src/InferTypesHandler.h
author František Kučera <franta-hg@frantovo.cz>
Sun, 23 May 2021 17:17:33 +0200
branchv_0
changeset 4 66c8a1783884
parent 1 da114916734b
child 9 215f62cc655f
permissions -rw-r--r--
do not keep Mode from previous relation
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
0
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     1
/**
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     2
 * Relational pipes
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     3
 * Copyright © 2021 František Kučera (Frantovo.cz, GlobalCode.info)
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     4
 *
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     5
 * This program is free software: you can redistribute it and/or modify
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     6
 * it under the terms of the GNU General Public License as published by
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     7
 * the Free Software Foundation, version 3 of the License.
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     8
 *
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     9
 * This program is distributed in the hope that it will be useful,
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    10
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    11
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    12
 * GNU General Public License for more details.
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    13
 *
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    14
 * You should have received a copy of the GNU General Public License
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    15
 * along with this program. If not, see <http://www.gnu.org/licenses/>.
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    16
 */
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    17
#pragma once
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    18
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    19
#include <memory>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    20
#include <string>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    21
#include <vector>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    22
#include <iostream>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    23
#include <sstream>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    24
#include <locale>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    25
#include <codecvt>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    26
#include <regex>
1
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    27
#include <stdexcept>
0
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    28
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    29
#include <relpipe/reader/typedefs.h>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    30
#include <relpipe/reader/TypeId.h>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    31
#include <relpipe/reader/handlers/RelationalReaderStringHandler.h>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    32
#include <relpipe/reader/handlers/AttributeMetadata.h>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    33
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    34
#include <relpipe/writer/Factory.h>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    35
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    36
#include <relpipe/cli/RelpipeCLIException.h>
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    37
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    38
#include "Configuration.h"
1
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    39
#include "Mode.h"
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    40
#include "DataMode.h"
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    41
#include "MetadataMode.h"
0
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    42
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    43
namespace relpipe {
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    44
namespace tr {
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    45
namespace infertypes {
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    46
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    47
using namespace std;
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    48
using namespace relpipe;
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    49
using namespace relpipe::reader;
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    50
using namespace relpipe::reader::handlers;
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    51
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    52
class InferTypesHandler : public RelationalReaderStringHandler {
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    53
private:
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    54
	shared_ptr<writer::RelationalWriter> relationalWriter;
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    55
	Configuration configuration;
1
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    56
	std::shared_ptr<Mode> currentMode;
0
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    57
1
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    58
	vector<writer::AttributeMetadata> toWriterMetadata(vector<AttributeMetadata> attributes) {
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    59
		// TODO: move to a reusable method (or use same metadata on both reader and writer side?)
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    60
		vector<writer::AttributeMetadata> writerMetadata;
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    61
		for (AttributeMetadata readerMetadata : attributes) {
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    62
			writerMetadata.push_back({readerMetadata.getAttributeName(), relationalWriter->toTypeId(readerMetadata.getTypeName())});
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    63
		}
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    64
		return writerMetadata;
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    65
	}
0
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    66
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    67
public:
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    68
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    69
	InferTypesHandler(shared_ptr<writer::RelationalWriter> relationalWriter, Configuration configuration) : relationalWriter(relationalWriter), configuration(configuration) {
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    70
	}
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    71
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    72
	void startRelation(string_t name, vector<AttributeMetadata> attributes) override {
1
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    73
		// TODO: move this logic to relpipe-lib-infertypes and share with certain modules like relpipe-in-csv
4
66c8a1783884 do not keep Mode from previous relation
František Kučera <franta-hg@frantovo.cz>
parents: 1
diff changeset
    74
		currentMode.reset();
1
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    75
		for (RelationConfiguration rc : configuration.relationConfigurations) {
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    76
			if (std::regex_match(name, rc.relationPattern)) {
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    77
				if (rc.mode == MODE::METADATA || rc.mode == MODE::AUTO && MetadataMode::willInfer(attributes)) currentMode.reset(new MetadataMode(relationalWriter));
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    78
				else if (rc.mode == MODE::DATA) currentMode.reset(new DataMode(relationalWriter));
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    79
				else throw std::logic_error("Unsupported mode: " + std::to_string((int) rc.mode));
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    80
				break;
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    81
			}
0
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    82
		}
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    83
1
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    84
		if (currentMode) currentMode->startRelation(name, attributes);
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    85
		else relationalWriter->startRelation(name, toWriterMetadata(attributes), true);
0
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    86
	}
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    87
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    88
	void attribute(const string_t& value) override {
1
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    89
		if (currentMode) currentMode->attribute(value);
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    90
		else relationalWriter->writeAttribute(value);
0
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    91
	}
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    92
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    93
	void endOfPipe() {
1
da114916734b multiple modes infrastructure
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    94
		currentMode.reset();
0
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    95
	}
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    96
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    97
};
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    98
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    99
}
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   100
}
5d1c556ff7db project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   101
}